COS 111-5
Reciprocal vegetation-flow feedbacks driving early-stage landscape evolution in a restored wet meadow

Thursday, August 13, 2015: 9:20 AM
344, Baltimore Convention Center
Danielle L Watts, Geography, University of California Berkeley, Berkeley, CA
Laurel L. Larsen, Geography, UC Berkeley, Berkeley, CA
Aayush Khurana, Geography, UC Berkeley, Berkeley, CA
Dorothy Merritts, Earth and Environment, Franklin and Marshall College, Lancaster, PA
Background/Question/Methods

Early stages of landscape development following disturbance provide a unique opportunity to delineate and understand ecogeomorphological feedback processes, as the diversity in plant ecomorphs is high and the potential for self-organization of landscape pattern strong. We used the opportunity of a stream restoration that reset its floodplain to 'initial conditions' to perform a suite of biophysical measurements designed to delineate the classes of feedback that influence landscape evolution in distinct ways. The Big Spring Run restoration (Lancaster, PA), completed in November 2011, involved removal of 15,000 t of legacy sediment from 1 km length valley bottom to expose a Holocene hydric layer and reestablish wet meadow hydrology and biota. By performing repeat biogeomorphic surveys within a study grid (16 x 28 m), we tested the hypothesis that distinct ecomorphs determine the persistence and location of channel and microtopographic features. The qualitatively distinct patch types surveyed included carpet-forming mat vegetation, tussock-forming vegetation, sparsely vegetated mudflats, benthic algal mats, mixed herbaceous communities, grasses, and clonal emergent vegetation. Within each sampling location, changes in vegetation community architecture, grain size distribution, critical shear stress for sediment entrainment, and topography were monitored over time, and flow resistance was measured. An overbank flow event that completely filled the floodplain provided an additional opportunity to measure vegetation-flow-sediment interactions. 

Results/Conclusions

In the three years of the study, distinct biogeomophic succession trajectories associated with different ecomorphs emerged: 1) Stabilization of sediment by benthic algae, followed by occupation by mat-forming vegetation and ultimately by grasses, 2) incipient new channel formation around clonal emergents or trees, which became secondary channels within 2 seasons, and 3) replacement of clonal emergents with grasses in areas not immediately adjacent to the channel or sustained by groundwater seeps. There has been a general trend of decreased ecomorph diversity, lowered diversity of flow across the floodplain, decreased flow resistance, and decreased critical shear stress for sediment entrainment. Overall, landscape dynamics in this restored wet meadow-floodplain are driven by deterministic succession trajectories but are influenced by stochastic elements such as the spatial distribution of groundwater seeps and the initial colonization of patch types that remain upright during high flow. Whereas hydrology might respond quickly to restoration change, biological, geological, and hydrological feedbacks might take years to stabilize.